INDEX

Explanations

references to negative impacts or consequences

oai_token-act-pair · gpt-3.5-turbo

references to the concept of "toll" and its associated impacts

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

itals

-0.81

 Correspond

-0.75

//[

-0.69

zsche

-0.66

 Sketch

-0.66

 Libre

-0.64

aple

-0.64

RGB

-0.62

ITY

-0.62

xual

-0.62

POSITIVE LOGITS

 toll

1.06

 booths

0.88

 Dickinson

0.82

 booth

0.79

 Toll

0.79

gate

0.78

sylvania

0.77

 plaza

0.74

 burden

0.72

Activations Density 0.024%

references to negative impacts or consequences

references to the concept of "toll" and its associated impacts

No Comments

No Known Activations