INDEX

Explanations

phrases related to political figures and government positions

oai_token-act-pair · gpt-3.5-turbo

references to political figures and their roles or actions

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 fibers

-0.70

 wells

-0.69

 mats

-0.69

 Scrib

-0.69

 selves

-0.69

 shards

-0.65

 electroly

-0.64

 ponds

-0.62

 intersection

-0.61

 seams

-0.61

POSITIVE LOGITS

 veto

0.97

ĳå£«

0.90

 appoint

0.88

 assassinated

0.84

 vetoed

0.84

 appointing

0.81

 Cabinet

0.79

 Downing

0.79

assad

0.78

 appointed

0.77

Activations Density 0.639%

phrases related to political figures and government positions

references to political figures and their roles or actions

No Comments

No Known Activations