INDEX

Explanations

references to individuals or entities being respected or highly regarded

oai_token-act-pair · gpt-3.5-turbo

references to individuals or entities that are regarded as respected or reputable

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

adra

-0.94

plet

-0.85

activated

-0.79

gradient

-0.77

Phones

-0.76

amaz

-0.75

occupied

-0.74

HUD

-0.74

plane

-0.73

agnetic

-0.72

POSITIVE LOGITS

 respected

0.89

 journalistic

0.85

 scholarly

0.79

 professionals

0.78

 reputable

0.78

 colleague

0.77

 venerable

0.76

 peer

0.76

 prestigious

0.74

 reputation

0.74

Activations Density 0.064%

references to individuals or entities being respected or highly regarded

references to individuals or entities that are regarded as respected or reputable

No Comments

No Known Activations