INDEX

Explanations

terms related to communication or organizations

oai_token-act-pair · gpt-3.5-turbo

terms related to community and communication

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.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.5.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

thur

-0.86

marked

-0.77

geons

-0.73

ctic

-0.72

teen

-0.71

geon

-0.69

wagen

-0.68

IVES

-0.68

 Flavoring

-0.68

 galleries

-0.67

POSITIVE LOGITS

mercial

0.73

angible

0.71

entary

0.69

Bei

0.68

ersion

0.68

ITED

0.68

rontal

0.67

achu

0.66

yang

0.65

ship

0.64

Activations Density 0.038%

terms related to communication or organizations

terms related to community and communication

No Comments

No Known Activations