INDEX

Explanations

words ending in '-led'

oai_token-act-pair · gpt-3.5-turbo

forms of the verb "to lead."

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.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.5.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

raviolet

-0.89

ĻĤ

-0.76

-+-+

-0.73

ħĭ

-0.70

ÑĢ

-0.69

Ħ¢

-0.67

ovi

-0.65

itu

-0.61

oln

-0.61

 teasp

-0.61

POSITIVE LOGITS

frog

0.82

ptin

0.78

puff

0.78

ome

0.77

cue

0.75

gling

0.75

ttes

0.74

bury

0.71

ogue

0.71

phrine

0.70

Activations Density 0.042%

words ending in '-led'

forms of the verb "to lead."

No Comments

No Known Activations