INDEX

Explanations

terms related to description or comparison of appearance

oai_token-act-pair · gpt-3.5-turbo

phrases indicating questioning or uncertainty about appearances and interpretations

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 3-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

rontal

-0.76

gart

-0.74

mop

-0.68

zar

-0.67

 Dickinson

-0.63

abad

-0.63

endo

-0.62

ournals

-0.61

ãĥĥ

-0.60

wash

-0.60

POSITIVE LOGITS

 entails

0.90

 fuss

0.76

really

0.75

pires

0.71

 entail

0.71

 boils

0.69

?:

0.69

?",

0.66

 REALLY

0.65

?,

0.64

Activations Density 0.678%

terms related to description or comparison of appearance

phrases indicating questioning or uncertainty about appearances and interpretations

No Comments

No Known Activations