INDEX

Explanations

mentions of specific job positions or roles at various organizations

oai_token-act-pair · gpt-3.5-turbo

the phrase "at" followed by various numbers or references to locations or times

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 11-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.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.11.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

unn

-0.65

boats

-0.64

reditary

-0.62

fill

-0.62

ulent

-0.61

Thumbnail

-0.61

oor

-0.61

ratulations

-0.58

oes

-0.58

comments

-0.58

POSITIVE LOGITS

 least

1.06

 Goldman

0.92

HHS

0.90

NYU

0.88

 UCLA

0.86

MIT

0.85

USC

0.85

 Pixar

0.85

abase

0.84

 Stanford

0.83

Activations Density 0.126%

mentions of specific job positions or roles at various organizations

the phrase "at" followed by various numbers or references to locations or times

No Comments

No Known Activations