INDEX

Explanations

references to self-identity

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 4-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_4/width_16k/average_l0_124

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.4.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

er

-0.84

palen

-0.67

erent

-0.66

位の

-0.65

Den

-0.65

TRI

-0.64

ER

-0.64

PreAuthorize

-0.63

-0.61

 czeko

-0.59

POSITIVE LOGITS

 myself

2.20

 yourself

2.11

myself

2.09

 herself

2.02

 ourselves

1.97

 himself

1.96

 Yourself

1.92

 Myself

1.87

herself

1.85

 Himself

1.84

Activations Density 0.058%

references to self-identity

No Comments

No Known Activations