INDEX

Explanations

the word "Clear", which may have various contexts but shows a strong association with this neuron

oai_token-act-pair · gpt-3.5-turbo

mentions of the word "Clear" and related variations

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 0-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.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.0.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

hob

-0.74

mim

-0.73

 dwar

-0.73

 imaginary

-0.70

 labou

-0.70

 homage

-0.66

 resp

-0.66

obos

-0.66

 repr

-0.65

pun

-0.65

POSITIVE LOGITS

 Clear

3.75

Clear

3.41

clear

2.36

 clear

1.52

CLE

1.40

 clearance

1.27

 clears

1.25

Cle

1.25

 Bright

1.18

 Plain

1.13

Activations Density 0.017%

the word "Clear", which may have various contexts but shows a strong association with this neuron

mentions of the word "Clear" and related variations

No Comments

No Known Activations