INDEX

Explanations

words related to government institutions, specifically the Parliament

oai_token-act-pair · gpt-3.5-turbo

mentions of legislative bodies or parliamentary contexts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.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

imilar

-0.68

ept

-0.65

ãģ¦

-0.65

-0.61

Complete

-0.60

resp

-0.60

False

-0.60

erm

-0.59

false

-0.59

 Decker

-0.57

POSITIVE LOGITS

 Parliament

1.15

liament

1.08

 parliament

1.05

arians

1.02

 Sabha

1.01

 ILCS

0.83

 Speaker

0.83

abulary

0.81

 Yard

0.79

 precinct

0.78

Activations Density 0.007%

words related to government institutions, specifically the Parliament

mentions of legislative bodies or parliamentary contexts

No Comments

No Known Activations