INDEX

Explanations

words related to choices or decision-making

oai_token-act-pair · gpt-3.5-turbo

references to the word "which" 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 @ 10-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.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

swick

-0.89

renheit

-0.79

ISTORY

-0.78

bart

-0.76

ibaba

-0.76

wn

-0.76

emp

-0.74

Ì¶

-0.74

shi

-0.73

yrinth

-0.73

POSITIVE LOGITS

 ones

1.18

 side

1.16

 direction

1.09

 wavelengths

1.03

 kinds

1.00

 aspects

0.99

 hemisphere

0.98

 parts

0.94

 subset

0.91

 facets

0.90

Activations Density 0.044%

words related to choices or decision-making

references to the word "which" in various contexts

No Comments

No Known Activations