INDEX

Explanations

phrases related to contributions or attributions

oai_token-act-pair · gpt-3.5-turbo

instances of the word "to" indicating contribution or causation

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

HUD

-0.70

Glass

-0.69

Champ

-0.69

CHAT

-0.68

IED

-0.66

STAT

-0.62

anned

-0.62

types

-0.62

framework

-0.61

rawler

-0.61

POSITIVE LOGITS

 compensate

0.81

 assist

0.78

 create

0.76

 amplify

0.76

 ensure

0.75

 counteract

0.75

 facilitate

0.75

 clot

0.73

 propagate

0.72

 accumulate

0.72

Activations Density 0.076%

phrases related to contributions or attributions

instances of the word "to" indicating contribution or causation

No Comments

No Known Activations