INDEX

Explanations

phrases related to composition or creation

oai_token-act-pair · gpt-3.5-turbo

instances of the word "made" and its variations 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 @ 8-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.8.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

assis

-0.64

 diarrhea

-0.59

":["

-0.58

 lows

-0.56

CLOSE

-0.55

Psy

-0.55

tnc

-0.55

 Regulatory

-0.54

 herbs

-0.54

ively

-0.53

POSITIVE LOGITS

 sure

0.85

Ò

0.84

 payable

0.83

solete

0.79

otype

0.76

 aware

0.74

URE

0.74

pez

0.71

otypes

0.71

 landfall

0.70

Activations Density 0.074%

phrases related to composition or creation

instances of the word "made" and its variations in various contexts

No Comments

No Known Activations