INDEX

Explanations

mentions of large or massive entities or concepts

oai_token-act-pair · gpt-3.5-turbo

mentions of size and scale 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 @ 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

speak

-0.77

mates

-0.76

 Modes

-0.73

 semantics

-0.72

ĪĴ

-0.72

 carbohyd

-0.71

Kind

-0.71

 Methods

-0.70

 Method

-0.70

 mates

-0.69

POSITIVE LOGITS

 footprint

1.15

 deficits

0.95

 conglomer

0.94

 chunk

0.93

 iceberg

0.92

 footprints

0.92

 deficit

0.90

 backlog

0.89

 reservoir

0.88

 influx

0.88

Activations Density 0.269%

mentions of large or massive entities or concepts

mentions of size and scale in various contexts

No Comments

No Known Activations