INDEX

Explanations

references to specific media outlets such as "The Daily Caller" and "Washington Examiner" in textual content

oai_token-act-pair · gpt-3.5-turbo

references to news outlets and publications

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 7-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.7.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.7.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

mble

-0.75

rero

-0.65

emb

-0.64

roman

-0.63

culated

-0.63

ogun

-0.63

leading

-0.62

ced

-0.62

oppable

-0.61

cul

-0.60

POSITIVE LOGITS

 Caller

0.94

NEWS

0.84

æ©Ł

0.71

 Nights

0.70

aires

0.68

Apps

0.67

iac

0.66

å¥³

0.64

oths

0.64

Ops

0.63

Activations Density 0.038%

references to specific media outlets such as "The Daily Caller" and "Washington Examiner" in textual content

references to news outlets and publications

No Comments

No Known Activations