INDEX

Explanations

number ranges listed in a specific format

oai_token-act-pair · gpt-3.5-turbo

phrases and expressions indicating ranges or comparisons

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 7-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.7.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.7.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

hemat

-0.67

aign

-0.65

nces

-0.59

 drafts

-0.58

intosh

-0.56

 depreciation

-0.55

inventory

-0.55

furt

-0.55

 hatched

-0.54

enance

-0.54

POSITIVE LOGITS

pless

1.07

ggles

1.06

pload

0.95

 insure

0.81

 fend

0.80

kus

0.79

asted

0.78

die

0.77

 fulfill

0.75

 ensure

0.75

Activations Density 0.059%

number ranges listed in a specific format

phrases and expressions indicating ranges or comparisons

No Comments

No Known Activations