INDEX

Explanations

phrases indicating accessibility to the public or various groups

oai_token-act-pair · gpt-3.5-turbo

phrases indicating openness and accessibility to the public

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

å¦

-0.72

Dur

-0.71

 Slug

-0.68

yan

-0.64

pring

-0.64

wise

-0.63

bowl

-0.63

arse

-0.62

chem

-0.61

urga

-0.61

POSITIVE LOGITS

 interpretation

0.96

 accommodate

0.87

 scrutiny

0.79

 newcomers

0.76

 outsiders

0.74

 Flavoring

0.74

 criticism

0.73

 explore

0.71

 experimentation

0.69

 exploration

0.67

Activations Density 0.122%

phrases indicating accessibility to the public or various groups

phrases indicating openness and accessibility to the public

No Comments

No Known Activations