INDEX

Explanations

references to political processes or events

oai_token-act-pair · gpt-3.5-turbo

mentions of referendums

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 0-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.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.0.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

ante

-0.76

atts

-0.75

Cap

-0.69

lin

-0.65

 overhead

-0.64

WARE

-0.63

azz

-0.63

 Aman

-0.61

aps

-0.61

 Wein

-0.61

POSITIVE LOGITS

 referendum

3.78

ndum

2.02

 refere

1.74

 ballot

1.59

 vote

1.53

 election

1.32

 Voters

1.31

election

1.30

Vote

1.29

 elections

1.28

Activations Density 0.010%

references to political processes or events

mentions of referendums

No Comments

No Known Activations