INDEX

Explanations

statements revealing important information or secrets

oai_token-act-pair · gpt-3.5-turbo

occurrences of the word "revelation."

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 0-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.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.0.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

onge

-0.76

eto

-0.68

abet

-0.66

ocial

-0.64

Gon

-0.63

ieg

-0.62

ymes

-0.61

 bart

-0.61

ickets

-0.61

 Supports

-0.59

POSITIVE LOGITS

 revelation

3.78

 revelations

3.00

 Revelations

1.82

 Revelation

1.49

 bombshell

1.45

 discovery

1.45

 discoveries

1.40

 disclosures

1.40

 confession

1.32

 reveal

1.31

Activations Density 0.024%

statements revealing important information or secrets

occurrences of the word "revelation."

No Comments

No Known Activations