INDEX

Explanations

locations, especially homes

oai_token-act-pair · gpt-3.5-turbo

mentions of "home" 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 @ 7-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.7.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.7.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

umbs

-0.76

andum

-0.71

umbers

-0.70

ACTED

-0.70

ãģĨ

-0.70

phies

-0.69

umper

-0.69

=-=-=-=-=-=-=-=-

-0.68

immers

-0.68

haps

-0.67

POSITIVE LOGITS

coming

0.97

pun

0.89

 turf

0.83

opathic

0.82

wife

0.81

opathy

0.79

nikov

0.78

 courtyard

0.78

 depot

0.78

opath

0.78

Activations Density 0.040%

locations, especially homes

mentions of "home" in various contexts

No Comments

No Known Activations