INDEX

Explanations

terms related to scientific experiments or studies

oai_token-act-pair · gpt-3.5-turbo

terms related to experimental processes or studies

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

veland

-0.84

andra

-0.78

utra

-0.75

ithing

-0.73

HCR

-0.71

atra

-0.71

olulu

-0.71

bery

-0.70

words

-0.70

criptions

-0.69

POSITIVE LOGITS

imental

0.94

 Prototype

0.74

 Experimental

0.73

ists

0.73

 experimental

0.72

 laboratory

0.72

 findings

0.70

 licence

0.69

 psychologists

0.69

 manip

0.69

Activations Density 0.015%

terms related to scientific experiments or studies

terms related to experimental processes or studies

No Comments

No Known Activations