INDEX

Explanations

negation terms indicating dissent or contradiction

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

New Auto-Interp

Top Features by Cosine Similarity

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

Configuration

google/gemma-scope-2b-pt-res/layer_2/width_16k/average_l0_141

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.2.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

Datuak

-0.87

SOUNDBITE

-0.82

StoryboardSegue

-0.81

ConstraintMaker

-0.81

 photolibrary

-0.73

 nakalista

-0.73

 noqa

-0.71

 ſever

-0.71

Autoritní

-0.70

 heiress

-0.70

POSITIVE LOGITS

not

0.98

Not

0.72

NOT

0.69

not

0.68

 likely

0.66

Not

0.65

NOT

0.56

 even

0.54

 typically

0.54

Activations Density 0.310%

negation terms indicating dissent or contradiction

No Comments

No Known Activations