INDEX

Explanations

Twitter usernames

oai_token-act-pair · gpt-3.5-turbo

proper nouns, particularly names and usernames

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 6-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.6.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.6.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

 substitutes

-0.74

theless

-0.73

 borne

-0.70

 ACTIONS

-0.70

é¾įå¥ĳå£«

-0.68

 substituted

-0.65

 resid

-0.65

 cured

-0.62

 press

-0.61

 uncertain

-0.60

POSITIVE LOGITS

uff

0.88

Magikarp

0.86

WithNo

0.85

Jr

0.84

trump

0.83

_.

0.83

Own

0.83

Whe

0.82

FT

0.82

td

0.81

Activations Density 0.079%

Twitter usernames

proper nouns, particularly names and usernames

No Comments

No Known Activations