INDEX

Explanations

mentions of physical proximity or closeness

oai_token-act-pair · gpt-3.5-turbo

variations of the word "close" and its context in phrases

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 11-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.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.11.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

AIN

-0.79

schild

-0.68

ilts

-0.68

Sov

-0.67

oat

-0.64

ains

-0.63

ICAN

-0.61

STE

-0.60

ants

-0.59

ims

-0.58

POSITIVE LOGITS

 proximity

1.05

 relatives

0.92

 knit

0.92

 quarters

0.89

 approximation

0.87

knit

0.85

 shave

0.84

 vicinity

0.83

 confid

0.81

enough

0.80

Activations Density 0.034%

mentions of physical proximity or closeness

variations of the word "close" and its context in phrases

No Comments

No Known Activations