INDEX

Explanations

words related to negative actions or characteristics, such as hate, murder, and suspicion

oai_token-act-pair · gpt-3.5-turbo

instances of the word "sn" or variations of it, likely indicating a focus on snarky or sarcastic commentary

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

heid

-0.87

EMENT

-0.76

xual

-0.72

WAYS

-0.64

mine

-0.63

 PowerPoint

-0.62

minus

-0.61

shire

-0.60

geist

-0.60

 Sacrament

-0.59

POSITIVE LOGITS

atching

1.19

obb

1.15

atches

1.13

agged

1.12

agging

1.11

appers

1.09

appy

1.08

apper

1.08

apping

1.07

ugg

1.06

Activations Density 0.009%

words related to negative actions or characteristics, such as hate, murder, and suspicion

instances of the word "sn" or variations of it, likely indicating a focus on snarky or sarcastic commentary

No Comments

No Known Activations