INDEX

Explanations

statements that critique the current societal narrative or highlight irony in discussions

oai_token-act-pair · gpt-4o-mini Triggered by @bot

expressions that negate or contradict claims, using patterns like "but [subject] [negation] [object]" to highlight contradictions or lack of evidence.

oai_token-act-pair · claude-4-5-haiku Triggered by @emiglarou

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B @ 31-gemmascope-res-16k

Configuration

google/gemma-scope-9b-pt-res/layer_31/width_16k/average_l0_114

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.31.hook_resid_post

Hook Layer

Architecture

jumprelu

Context Size

1,024

Dataset

monology/pile-uncopyrighted

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

ródło

-0.40

 geheel

-0.38

いきなり

-0.37

XmlAccessorType

-0.36

 nécessairement

-0.35

ępnie

-0.34

ByVersion

-0.34

 скорее

-0.34

 kaarangay

-0.34

aarrggbb

-0.34

POSITIVE LOGITS

 slightest

1.04

 even

0.79

 ounce

0.75

IntoConstraints

0.74

哪怕

0.73

 choć

0.73

 iota

0.72

 siquiera

0.71

 хоть

0.71

 moindre

0.69

Activations Density 0.361%

statements that critique the current societal narrative or highlight irony in discussions

expressions that negate or contradict claims, using patterns like "but [subject] [negation] [object]" to highlight contradictions or lack of evidence.

No Comments

No Known Activations

statements that critique the current societal narrative or highlight irony in discussions

expressions that negate or contradict claims, using patterns like "but [subject] [negation] [object]" to highlight contradictions or lack of evidence.

No Comments

No Known Activations