INDEX

Explanations

expressions of approval or disapproval towards political figures or actions

oai_token-act-pair · gpt-3.5-turbo

phrases related to public approval and disapproval ratings

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 7-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.7.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.7.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

orage

-0.66

GV

-0.64

 fixme

-0.64

 Roche

-0.62

udic

-0.61

mith

-0.61

UNCH

-0.61

AAF

-0.60

 Codex

-0.59

esis

-0.59

POSITIVE LOGITS

 rating

1.00

 ratings

0.99

rating

0.92

renheit

0.85

Rating

0.83

rated

0.81

 Gallup

0.76

 polarization

0.76

 disapprove

0.75

watching

0.75

Activations Density 0.125%

expressions of approval or disapproval towards political figures or actions

phrases related to public approval and disapproval ratings

No Comments

No Known Activations