INDEX

Explanations

phrases or expressions related to common, regular, or expected things or events

oai_token-act-pair · gpt-3.5-turbo

references to the concept of the "usual" 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 @ 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

kamp

-0.86

mentation

-0.78

possibly

-0.77

mented

-0.77

OPE

-0.74

 finally

-0.68

acus

-0.68

lang

-0.68

better

-0.68

another

-0.68

POSITIVE LOGITS

 suspects

1.29

 assortment

1.10

 disclaimer

1.03

 disclaim

0.93

 fare

0.92

 caveats

0.91

 boiler

0.83

 pitfalls

0.83

 tropes

0.82

 rundown

0.79

Activations Density 0.093%

phrases or expressions related to common, regular, or expected things or events

references to the concept of the "usual" in various contexts

No Comments

No Known Activations