INDEX

Explanations

the word "no" followed by a number indicating the strength of the activation

oai_token-act-pair · gpt-3.5-turbo

negations or expressions of absence

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.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

RAFT

-0.85

adobe

-0.77

Quote

-0.71

rex

-0.71

lycer

-0.70

endar

-0.70

Untitled

-0.69

assies

-0.68

iership

-0.67

north

-0.66

POSITIVE LOGITS

xious

1.13

 longer

1.09

 matter

1.01

 doubt

0.97

etheless

0.94

except

0.90

 conceivable

0.88

oooo

0.85

 indication

0.84

 clue

0.84

Activations Density 0.100%

the word "no" followed by a number indicating the strength of the activation

negations or expressions of absence

No Comments

No Known Activations