INDEX

Explanations

words related to regular and consistent actions or activities

oai_token-act-pair · gpt-3.5-turbo

instances of the word "routinely" in various contexts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 2-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.2.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.2.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

lad

-0.73

soc

-0.72

 Millennium

-0.72

jri

-0.68

elo

-0.63

Kah

-0.63

 Chains

-0.62

 Intent

-0.61

 Emir

-0.61

 Horizons

-0.61

POSITIVE LOGITS

theless

0.92

 pract

0.85

entimes

0.83

 encountered

0.81

 subjected

0.78

ãĤ©

0.78

 complained

0.78

 disreg

0.77

 employed

0.76

©¶æ

0.76

Activations Density 0.008%

words related to regular and consistent actions or activities

instances of the word "routinely" in various contexts

No Comments

No Known Activations