INDEX

Explanations

phrases related to official documents or communication in a professional or organized setting

oai_token-act-pair · gpt-3.5-turbo

occurrences of the word "letter"

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

rolet

-0.75

negie

-0.72

rises

-0.66

 Occupations

-0.63

 Tune

-0.62

 Bronze

-0.62

issan

-0.60

 CCTV

-0.60

 STATS

-0.59

ptions

-0.58

POSITIVE LOGITS

Letter

1.04

 addressed

0.99

 letter

0.97

velop

0.92

letter

0.91

 sent

0.89

 letters

0.88

 correspondence

0.86

 mailed

0.86

 inbox

0.86

Activations Density 0.045%

phrases related to official documents or communication in a professional or organized setting

occurrences of the word "letter"

No Comments

No Known Activations