INDEX

Explanations

connections between actions and outcomes

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

following "for," "in," and "on"

np_acts-logits-general · gemini-2.0-flash

preposition + common phrase word

np_acts-logits-general · gemini-2.5-flash-lite

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 22-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_22/width_16k/average_l0_72

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.22.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

rawDesc

-0.90

 RIPRODUZIONE

-0.81

 préfé

-0.71

 getArguments

-0.70

twimg

-0.66

клопе

-0.64

 quæ

-0.64

feira

-0.62

DispatchToProps

-0.62

 irré

-0.61

POSITIVE LOGITS

expandindo

0.68

rungsseite

0.56

)]

0.55

#+#

0.53

tvguidetime

0.53

tedrale

0.51

]`

0.51

 fucking

0.51

HtmlAttribute

0.51

'],

0.50

Activations Density 1.500%

connections between actions and outcomes

following "for," "in," and "on"

preposition + common phrase word

No Comments

No Known Activations

connections between actions and outcomes

following "for," "in," and "on"

preposition + common phrase word

No Comments

No Known Activations