INDEX

Explanations

words related to conflicts or disputes, especially those involving physical altercations

oai_token-act-pair · gpt-3.5-turbo

references to conflicts or violent altercations

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

izons

-0.75

agles

-0.66

toe

-0.66

travel

-0.64

sonian

-0.64

rolog

-0.62

ignty

-0.62

 sails

-0.61

 Editors

-0.61

TPP

-0.60

POSITIVE LOGITS

 spree

1.17

 involving

0.96

 altercation

0.89

 perpetrated

0.86

 occurred

0.78

 burglary

0.78

 Rampage

0.76

 ensued

0.76

 rampage

0.75

 aggravated

0.74

Activations Density 0.159%

words related to conflicts or disputes, especially those involving physical altercations

references to conflicts or violent altercations

No Comments

No Known Activations