INDEX

Explanations

questions

oai_token-act-pair · gpt-3.5-turbo

question marks and interrogative punctuation

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

 Swanson

-0.77

shaw

-0.69

ality

-0.64

 slam

-0.63

 Nadu

-0.62

apers

-0.60

 hydra

-0.60

ersen

-0.59

 proced

-0.59

 laced

-0.58

POSITIVE LOGITS

ï¿½ï¿½ï¿½ï¿½

1.06

utm

0.93

.?

0.89

?:

0.89

ï¸ı

0.77

crop

0.75

rencies

0.73

hhhh

0.73

ãĥ¤

0.72

IQ

0.70

Activations Density 0.045%

questions

question marks and interrogative punctuation

No Comments

No Known Activations