INDEX

Explanations

words related to giving a new perspective or interpretation

oai_token-act-pair · gpt-3.5-turbo

instances of the word "spin" 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 @ 4-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.4.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.4.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

avis

-0.72

 Admir

-0.71

 Commodore

-0.67

ecause

-0.65

inances

-0.64

inez

-0.63

 Scores

-0.62

ĨĴ

-0.62

enance

-0.62

 Defenders

-0.59

POSITIVE LOGITS

ners

1.42

 spin

1.09

spin

0.96

kered

0.91

eless

0.91

wheel

0.86

cher

0.82

rot

0.82

ball

0.80

ned

0.80

Activations Density 0.009%

words related to giving a new perspective or interpretation

instances of the word "spin" in various contexts

No Comments

No Known Activations