INDEX

Explanations

time-related phrases such as durations and specific time spans

oai_token-act-pair · gpt-3.5-turbo

references to time, specifically durations of years and months

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

arton

-0.63

XII

-0.62

 Hearts

-0.61

 Sixers

-0.60

Phi

-0.60

inite

-0.59

Haw

-0.58

Sin

-0.58

alg

-0.58

Dru

-0.57

POSITIVE LOGITS

ago

1.08

catentry

0.75

Ago

0.72

izo

0.71

vez

0.69

 elapsed

0.66

 WARRANT

0.64

 gestation

0.64

depending

0.64

>]

0.64

Activations Density 0.136%

time-related phrases such as durations and specific time spans

references to time, specifically durations of years and months

No Comments

No Known Activations