INDEX

Explanations

the word 'kel' with different variations in surrounding contexts

oai_token-act-pair · gpt-3.5-turbo

proper nouns related to individuals and their actions

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 3-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.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

Interstitial

-0.75

IDENT

-0.72

displayText

-0.72

England

-0.68

 Parables

-0.66

Critical

-0.66

Reply

-0.65

LEASE

-0.64

EED

-0.64

reads

-0.63

POSITIVE LOGITS

stein

1.22

bach

1.03

sten

1.01

anguage

1.00

mann

0.96

idon

0.93

kel

0.92

mont

0.88

jriwal

0.86

kie

0.86

Activations Density 0.006%

the word 'kel' with different variations in surrounding contexts

proper nouns related to individuals and their actions

No Comments

No Known Activations