INDEX

Explanations

words related to criminal activities or law enforcement

oai_token-act-pair · gpt-3.5-turbo

the word "der" 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 @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.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.5.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

ainted

-0.63

ERC

-0.60

els

-0.59

 Babe

-0.59

ELS

-0.59

Fey

-0.57

 Decay

-0.56

oufl

-0.56

Sri

-0.55

zza

-0.55

POSITIVE LOGITS

dash

1.11

iving

0.89

hoe

0.83

iver

0.83

rors

0.83

isively

0.83

netic

0.82

minster

0.82

ror

0.82

cair

0.81

Activations Density 0.018%

words related to criminal activities or law enforcement

the word "der" in various contexts

No Comments

No Known Activations