INDEX

Explanations

phrases related to difficulty or complexity

oai_token-act-pair · gpt-3.5-turbo

statements about the existence or condition of things

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 11-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.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.11.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

dylib

-0.78

 throats

-0.78

luaj

-0.70

enium

-0.70

rower

-0.68

ãĤ£

-0.66

rul

-0.66

ayn

-0.66

ipient

-0.65

iat

-0.62

POSITIVE LOGITS

 tricky

1.13

 akin

1.12

 fraught

1.02

 risky

1.01

 daunting

1.01

 futile

1.00

 easier

0.99

 difficult

0.99

 impossible

0.97

 exhausting

0.97

Activations Density 0.166%

phrases related to difficulty or complexity

statements about the existence or condition of things

No Comments

No Known Activations