INDEX

Explanations

words related to influence, guidance, and conditioning

oai_token-act-pair · gpt-3.5-turbo

pronouns and phrases related to guidance and influence

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

tein

-0.69

 Shotgun

-0.59

 roundup

-0.58

 Pastebin

-0.57

 Federation

-0.57

 understatement

-0.56

mean

-0.55

 Radar

-0.54

Maw

-0.54

 Confeder

-0.54

POSITIVE LOGITS

selves

1.05

 accordingly

0.93

self

0.84

 psychologically

0.83

 anew

0.83

atically

0.82

heit

0.80

onto

0.78

 spiritually

0.78

into

0.77

Activations Density 0.374%

words related to influence, guidance, and conditioning

pronouns and phrases related to guidance and influence

No Comments

No Known Activations