INDEX

Explanations

phrases indicating a time period expressed with the word "later"

oai_token-act-pair · gpt-3.5-turbo

references to time intervals or delays

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

owship

-0.79

hab

-0.78

rament

-0.63

heet

-0.60

TOD

-0.60

kef

-0.59

ende

-0.58

lua

-0.57

tw

-0.57

Ĥª

-0.57

POSITIVE LOGITS

 arrivals

0.72

aneously

0.72

forth

0.70

 graduation

0.68

 Pats

0.68

ettings

0.66

 earthqu

0.66

noon

0.65

 succumb

0.65

 regretted

0.65

Activations Density 0.026%

phrases indicating a time period expressed with the word "later"

references to time intervals or delays

No Comments

No Known Activations

phrases indicating a time period expressed with the word "later"

references to time intervals or delays

No Comments

No Known Activations