INDEX

Explanations

proper nouns related to politics and personalities

oai_token-act-pair · gpt-3.5-turbo

specific names and identifiers of individuals or entities

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 11-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.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.11.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

REDACTED

-0.74

 dearly

-0.68

 IMAGES

-0.65

iliary

-0.62

 stewards

-0.61

llor

-0.60

govtrack

-0.60

 stabilization

-0.59

minecraft

-0.58

 notch

-0.56

POSITIVE LOGITS

abeth

0.96

abet

0.92

steen

0.88

icio

0.85

ãĤ¶

0.83

ande

0.82

OY

0.80

apo

0.78

ukes

0.76

ador

0.75

Activations Density 0.147%

proper nouns related to politics and personalities

specific names and identifiers of individuals or entities

No Comments

No Known Activations