INDEX

Explanations

verbs related to actions or processes

oai_token-act-pair · gpt-3.5-turbo

actions or verbs that suggest significant impact or change

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

iscons

-0.72

hatt

-0.72

fty

-0.71

 reluct

-0.68

 confir

-0.67

imen

-0.66

pard

-0.66

aniel

-0.65

 cabinet

-0.64

odore

-0.64

POSITIVE LOGITS

 GMOs

0.67

 themselves

0.65

verages

0.65

ides

0.62

 Powered

0.59

 miscar

0.59

 otherwise

0.58

LW

0.58

 Forsaken

0.58

rament

0.58

Activations Density 0.240%

verbs related to actions or processes

actions or verbs that suggest significant impact or change

No Comments

No Known Activations