INDEX

Explanations

phrases related to political figures, actions, and opinions

oai_token-act-pair · gpt-3.5-turbo

instances of political action and conflict

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

!--

-0.71

ruary

-0.69

 Paste

-0.63

theless

-0.63

vertisements

-0.60

 "$:/

-0.60

 attm

-0.60

mania

-0.57

usercontent

-0.55

pmwiki

-0.55

POSITIVE LOGITS

 Kavanaugh

0.96

 Rosenstein

0.82

 Dems

0.79

 opioid

0.70

ħĭ

0.68

 Pelosi

0.65

İ

0.64

Ford

0.64

 Bannon

0.64

Ħ¢

0.63

Activations Density 0.045%

phrases related to political figures, actions, and opinions

instances of political action and conflict

No Comments

No Known Activations