INDEX

Explanations

names of individuals or organizations

oai_token-act-pair · gpt-3.5-turbo

phrases or terms related to advertisements and commercial content

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

Lex

-0.79

 Paran

-0.73

 Corinth

-0.70

steen

-0.70

 Aston

-0.68

 Winn

-0.68

 Lexington

-0.67

 Lionel

-0.67

anni

-0.66

 embell

-0.65

POSITIVE LOGITS

Į

0.91

ocy

0.82

é¾įå

0.80

yna

0.77

cffff

0.76

uh

0.76

Shift

0.71

ately

0.71

ãĥĥãĥĪ

0.71

aez

0.70

Activations Density 0.373%

names of individuals or organizations

phrases or terms related to advertisements and commercial content

No Comments

No Known Activations