INDEX

Explanations

phrases indicating unnecessary information or obvious statements

oai_token-act-pair · gpt-3.5-turbo

spoken expressions or phrases suggesting a need for clarification

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

anmar

-0.76

ahime

-0.71

 Flavoring

-0.69

iology

-0.66

assadors

-0.63

luaj

-0.60

 mediated

-0.59

ipolar

-0.59

chwitz

-0.59

 coordination

-0.57

POSITIVE LOGITS

 conclud

0.74

WER

0.68

CTV

0.67

 caveats

0.65

 Hers

0.65

 Investigator

0.63

 disclaim

0.62

lege

0.62

 fert

0.61

 Patriot

0.61

Activations Density 0.141%

phrases indicating unnecessary information or obvious statements

spoken expressions or phrases suggesting a need for clarification

No Comments

No Known Activations