INDEX

Explanations

locations or objects situated in a high position

oai_token-act-pair · gpt-3.5-turbo

phrases indicating a position or location in relation to other objects

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.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.5.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

len

-0.81

gm

-0.79

nel

-0.73

ties

-0.70

Ð´

-0.70

nery

-0.69

nc

-0.68

ners

-0.68

arians

-0.66

zel

-0.66

POSITIVE LOGITS

ĺħ

0.92

 atop

0.89

owered

0.76

ĪĴ

0.74

ļé

0.74

hap

0.73

eous

0.73

İĭ

0.72

avorite

0.71

htaking

0.71

Activations Density 0.030%

locations or objects situated in a high position

phrases indicating a position or location in relation to other objects

No Comments

No Known Activations