INDEX

Explanations

words related to written communication or documents

oai_token-act-pair · gpt-3.5-turbo

references to specific letters or correspondence in a variety of contexts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 10-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.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

rises

-0.71

orsi

-0.68

rolet

-0.66

eger

-0.66

issan

-0.66

alon

-0.65

 CCTV

-0.63

 STATS

-0.62

ESCO

-0.61

 Occupations

-0.60

POSITIVE LOGITS

press

1.07

boxing

0.96

Letter

0.93

letter

0.93

box

0.91

boxes

0.87

 addressed

0.86

head

0.85

velop

0.85

sworth

0.84

Activations Density 0.032%

words related to written communication or documents

references to specific letters or correspondence in a variety of contexts

No Comments

No Known Activations