INDEX

Explanations

questions related to seeking information or guidance on various topics

oai_token-act-pair · gpt-3.5-turbo

questions and phrases related to obtaining or understanding information

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

hoff

-0.69

gru

-0.62

pict

-0.62

foundland

-0.60

posts

-0.60

artifacts

-0.59

Ń·

-0.59

court

-0.58

 Frazier

-0.58

arov

-0.57

POSITIVE LOGITS

?:

0.75

emis

0.68

uate

0.67

oint

0.67

?]

0.67

owe

0.67

0.65

?),

0.64

riet

0.62

AME

0.61

Activations Density 0.139%

questions related to seeking information or guidance on various topics

questions and phrases related to obtaining or understanding information

No Comments

No Known Activations