INDEX

Explanations

mentions of drag-related concepts or actions

oai_token-act-pair · gpt-3.5-turbo

occurrences of the term "drag" and its variations in different 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

vironment

-0.70

AVG

-0.67

 Kubrick

-0.66

 Blueprint

-0.64

theless

-0.64

ership

-0.60

zbek

-0.59

places

-0.58

leck

-0.57

etheless

-0.56

POSITIVE LOGITS

oon

1.18

ging

1.07

 queens

1.04

ged

0.99

net

0.97

strip

0.94

gin

0.91

gy

0.86

gery

0.85

gers

0.84

Activations Density 0.063%

mentions of drag-related concepts or actions

occurrences of the term "drag" and its variations in different contexts

No Comments

No Known Activations