INDEX

Explanations

words related to political affiliations and leanings

oai_token-act-pair · gpt-3.5-turbo

references to political affiliations and terms related to independent voters

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 3-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.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

ãĥ¥

-0.77

itual

-0.76

advertising

-0.73

pter

-0.70

MER

-0.69

ãĤ¨ãĥ«

-0.69

fare

-0.68

ASED

-0.67

 bacter

-0.67

geon

-0.67

POSITIVE LOGITS

 independents

0.88

ecided

0.84

 strugg

0.74

essor

0.74

 leaning

0.74

 voters

0.73

ratulations

0.73

clair

0.72

 polled

0.72

Cath

0.71

Activations Density 0.013%

words related to political affiliations and leanings

references to political affiliations and terms related to independent voters

No Comments

No Known Activations