INDEX

Explanations

references to "The Associated Press."

oai_token-act-pair · gpt-3.5-turbo

mentions of the Associated Press

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

 Luthor

-0.73

gone

-0.71

sbm

-0.71

milo

-0.70

lycer

-0.68

ouses

-0.67

 unfocusedRange

-0.67

 capitalists

-0.67

 graphene

-0.66

 Gerrard

-0.65

POSITIVE LOGITS

PLIED

0.84

 Corpus

0.82

 Agency

0.81

wire

0.79

eal

0.75

 Release

0.74

 Herald

0.74

 dispatch

0.73

 Examiner

0.71

 Coverage

0.70

Activations Density 0.011%

references to "The Associated Press."

mentions of the Associated Press

No Comments

No Known Activations