INDEX

Explanations

terms related to military and politics

oai_token-act-pair · gpt-3.5-turbo

terms related to humanitarian issues and activism

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

enery

-0.81

 Alloy

-0.74

urer

-0.69

URES

-0.67

 Yelp

-0.65

ĸļ

-0.64

FACE

-0.63

 Thumbnails

-0.63

uring

-0.63

 photos

-0.61

POSITIVE LOGITS

ians

1.00

ters

0.99

ilateral

0.87

ilty

0.87

itar

0.83

iak

0.82

iac

0.79

tered

0.78

ypes

0.78

ic

0.77

Activations Density 0.027%

terms related to military and politics

terms related to humanitarian issues and activism

No Comments

No Known Activations