INDEX

Explanations

terms related to legal or official titles and acts

oai_token-act-pair · gpt-3.5-turbo

terms related to official organizations, actions, or legislation

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 6-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.6.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.6.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

theless

-1.03

 altogether

-0.71

 elves

-0.71

 turkey

-0.69

err

-0.68

 oneself

-0.68

 sake

-0.68

 overhead

-0.67

 dipping

-0.67

 goblins

-0.67

POSITIVE LOGITS

iciency

0.97

istries

0.96

iliation

0.93

enses

0.92

 Registry

0.92

itions

0.91

ilit

0.91

uctions

0.91

entials

0.91

ocation

0.91

Activations Density 0.238%

terms related to legal or official titles and acts

terms related to official organizations, actions, or legislation

No Comments

No Known Activations