INDEX

Explanations

phrases or sentences indicating someone was caught doing something, often with negative connotations

oai_token-act-pair · gpt-3.5-turbo

instances of the word "caught" in various contexts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 3-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.hook_resid_pre

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

24,576

Data Type

torch.float32

Hook Point

blocks.3.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 pronoun

-0.77

 pronouns

-0.75

 helicop

-0.70

umably

-0.67

 livest

-0.67

edom

-0.66

 reluct

-0.66

anwhile

-0.66

 persuasion

-0.65

hov

-0.65

POSITIVE LOGITS

ãĤµ

0.85

 caught

0.75

Rai

0.75

ãĤ¤

0.74

 glimps

0.72

Frames

0.72

phrase

0.72

aught

0.72

 Luffy

0.70

netflix

0.70

Activations Density 0.016%

phrases or sentences indicating someone was caught doing something, often with negative connotations

instances of the word "caught" in various contexts

No Comments

No Known Activations