INDEX

Explanations

the word "keys" along with their respective strength of association

oai_token-act-pair · gpt-3.5-turbo

repeated mentions of "keys" in various contexts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 2-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.2.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.2.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

ATIONAL

-0.77

 Colleges

-0.70

Marie

-0.69

ews

-0.69

URRENT

-0.66

 Wolver

-0.66

 PHOTO

-0.65

 Vide

-0.65

renheit

-0.64

 Recreation

-0.63

POSITIVE LOGITS

 keys

1.13

*/(

1.09

stroke

1.09

keys

1.05

stem

0.93

chain

0.89

Keys

0.89

holes

0.87

notes

0.87

bind

0.86

Activations Density 0.014%

the word "keys" along with their respective strength of association

repeated mentions of "keys" in various contexts

No Comments

No Known Activations