INDEX

Explanations

the word "one" followed by a number indicating intensity

oai_token-act-pair · gpt-3.5-turbo

instances of the word "one."

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

older

-0.82

mson

-0.75

apons

-0.70

aeda

-0.70

ador

-0.70

hess

-0.67

enery

-0.66

estones

-0.65

inders

-0.64

cases

-0.63

POSITIVE LOGITS

 sided

1.00

 dimensional

0.98

 Hundred

0.92

 minute

0.91

 hundred

0.87

 Minute

0.81

day

0.81

 Piece

0.81

 hour

0.79

handedly

0.79

Activations Density 0.110%

the word "one" followed by a number indicating intensity

instances of the word "one."

No Comments

No Known Activations