INDEX

Explanations

references to legal or official statuses

oai_token-act-pair · gpt-3.5-turbo

terms related to legal status and political protection

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 10-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

ndra

-0.69

 Tale

-0.69

ofi

-0.68

rame

-0.65

assian

-0.63

rican

-0.62

ramid

-0.62

orah

-0.61

erer

-0.60

ollar

-0.59

POSITIVE LOGITS

 again

0.91

cards

0.85

 credentials

0.81

 anew

0.78

DERR

0.77

 himself

0.74

 twice

0.74

 reinforcements

0.74

 sufficient

0.74

for

0.73

Activations Density 0.342%

references to legal or official statuses

terms related to legal status and political protection

No Comments

No Known Activations