INDEX

Explanations

verbal expressions indicating outcomes or revelations

oai_token-act-pair · gpt-3.5-turbo

phrases that include the word "turn" in various forms

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 9-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.9.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.9.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

 resembled

-0.59

nces

-0.58

 resembles

-0.56

 idiots

-0.55

lihood

-0.53

 dism

-0.52

 resemble

-0.51

 starved

-0.50

 everyday

-0.50

 anytime

-0.50

POSITIVE LOGITS

llor

0.71

erc

0.66

aran

0.65

erg

0.64

rue

0.64

chn

0.64

ffe

0.64

ere

0.63

hoff

0.59

ede

0.59

Activations Density 0.356%

verbal expressions indicating outcomes or revelations

phrases that include the word "turn" in various forms

No Comments

No Known Activations