INDEX

Explanations

words related to a specific name or topic ("ler")

oai_token-act-pair · gpt-3.5-turbo

references to individuals, particularly last names or surnames

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.8.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

liest

-0.69

ADRA

-0.65

UAL

-0.63

SIM

-0.62

popular

-0.60

ERN

-0.55

ouls

-0.54

Latin

-0.54

kok

-0.54

 solid

-0.54

POSITIVE LOGITS

oad

1.01

istics

0.94

sonian

0.90

ogue

0.89

bach

0.88

iffe

0.88

gren

0.87

kamp

0.86

baugh

0.86

oyal

0.85

Activations Density 0.058%

words related to a specific name or topic ("ler")

references to individuals, particularly last names or surnames

No Comments

No Known Activations