INDEX

Explanations

references to specific routes or locations

oai_token-act-pair · gpt-3.5-turbo

mentions of specific routes

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 10-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.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

uster

-0.90

iaries

-0.82

luaj

-0.78

endi

-0.75

oppable

-0.71

usters

-0.70

oice

-0.69

ilts

-0.69

amily

-0.68

bara

-0.68

POSITIVE LOGITS

finding

1.01

 routes

1.00

Route

0.90

 route

0.88

finder

0.88

route

0.86

WAYS

0.82

 routing

0.79

 Navigation

0.78

 itiner

0.76

Activations Density 0.029%

references to specific routes or locations

mentions of specific routes

No Comments

No Known Activations