INDEX

Explanations

political and government-related terms and events

oai_token-act-pair · gpt-3.5-turbo

occurrences of significant political events or decisions

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

)).

-0.60

uild

-0.58

jew

-0.57

lection

-0.56

fur

-0.56

tty

-0.54

]).

-0.53

rity

-0.53

urses

-0.52

cture

-0.51

POSITIVE LOGITS

razil

0.60

slaught

0.55

utterstock

0.55

gado

0.55

MFT

0.50

 weakening

0.50

FG

0.49

odka

0.49

 Metatron

0.49

 Liber

0.48

Activations Density 1.784%

political and government-related terms and events

occurrences of significant political events or decisions

No Comments

No Known Activations