INDEX

Explanations

words related to linguistic and computational terms, specifically words ending in '-ing', '-ed', '-ly', '-er', or '-s'

oai_token-act-pair · gpt-3.5-turbo

words related to linguistic forms and suffixes

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 9-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.9.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.9.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

KP

-0.70

schild

-0.69

iewicz

-0.67

 Hayden

-0.65

 Sutherland

-0.63

 Immunity

-0.63

 Helsinki

-0.62

Luk

-0.61

TAM

-0.60

 Elves

-0.60

POSITIVE LOGITS

mble

1.00

cean

0.92

moil

0.87

uin

0.85

gment

0.82

renheit

0.81

rill

0.80

mph

0.80

odox

0.79

ased

0.78

Activations Density 0.168%

words related to linguistic and computational terms, specifically words ending in '-ing', '-ed', '-ly', '-er', or '-s'

words related to linguistic forms and suffixes

No Comments

No Known Activations