INDEX

Explanations

statements where something is being asserted or claimed

oai_token-act-pair · gpt-3.5-turbo

statements that involve claims or assertions made by individuals or organizations

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.8.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

Joined

-0.70

 CrossRef

-0.67

 Graveyard

-0.64

 Springer

-0.62

 unfocusedRange

-0.62

Suddenly

-0.58

 goose

-0.58

 graveyard

-0.57

igslist

-0.56

 funeral

-0.56

POSITIVE LOGITS

 innocence

0.98

 nothing

0.77

absolutely

0.76

 abide

0.76

ItemImage

0.75

alty

0.74

IFIED

0.72

 neutrality

0.72

 abiding

0.69

 harmless

0.68

Activations Density 0.405%

statements where something is being asserted or claimed

statements that involve claims or assertions made by individuals or organizations

No Comments

No Known Activations