INDEX

Explanations

locations or references to physical areas marked as 'stretch'

oai_token-act-pair · gpt-3.5-turbo

the repeated term "stretch" in various contexts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 3-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.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

eer

-0.89

nect

-0.87

eers

-0.81

*/(

-0.78

obs

-0.78

ellery

-0.74

ept

-0.72

iatric

-0.70

vernment

-0.68

abad

-0.68

POSITIVE LOGITS

 stretch

0.94

Stretch

0.87

 stretches

0.84

 stretched

0.83

 Afric

0.78

retch

0.73

aways

0.71

lim

0.70

opian

0.70

 stretching

0.68

Activations Density 0.007%

locations or references to physical areas marked as 'stretch'

the repeated term "stretch" in various contexts

No Comments

No Known Activations