INDEX

Explanations

phrases indicating a sense of limitation or restriction

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res_fs24576-jb

Configuration

jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_24576

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

 mull

-0.65

estamp

-0.63

Dru

-0.63

uscript

-0.59

 extensively

-0.55

 corpus

-0.55

 Lomb

-0.55

cius

-0.54

atana

-0.54

ouncing

-0.53

POSITIVE LOGITS

 worsened

0.66

 exacerbate

0.64

ayed

0.64

aires

0.64

ĵĺ

0.63

¥µ

0.63

ICES

0.61

 fools

0.61

CHAT

0.60

ĨĴ

0.60

Activations Density 0.092%

phrases indicating a sense of limitation or restriction

No Comments

No Known Activations