INDEX

Explanations

terms related to actions like printing, sharing, and emailing content

oai_token-act-pair · gpt-3.5-turbo

references to print-related actions or features

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 7-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.7.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.7.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

atos

-0.58

 Myster

-0.58

ales

-0.58

hips

-0.57

rower

-0.55

itance

-0.55

isco

-0.54

itational

-0.54

opoulos

-0.54

eson

-0.53

POSITIVE LOGITS

icut

0.73

ãĥĥãĥī

0.71

 Thumbnails

0.70

Ô

0.68

imedia

0.64

Blog

0.64

lihood

0.59

pleting

0.56

VID

0.56

å¦

0.56

Activations Density 0.253%

terms related to actions like printing, sharing, and emailing content

references to print-related actions or features

No Comments

No Known Activations