INDEX

Explanations

queries that begin with "How do"

oai_token-act-pair · gpt-3.5-turbo

sentences that pose questions about actions or inquiries

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

 Immunity

-0.72

 Handling

-0.68

board

-0.66

Reviewer

-0.65

workshop

-0.65

bane

-0.64

 ridden

-0.64

Jar

-0.64

cised

-0.63

mares

-0.62

POSITIVE LOGITS

omsday

1.12

ppel

0.84

herty

0.77

onga

0.73

zens

0.73

?]

0.72

ctr

0.72

kson

0.69

orman

0.69

 impressions

0.68

Activations Density 0.026%

queries that begin with "How do"

sentences that pose questions about actions or inquiries

No Comments

No Known Activations