INDEX

Explanations

government-related terms like titles, positions, and institutions

oai_token-act-pair · gpt-3.5-turbo

references to governmental or state-related entities and processes

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.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.5.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

ï¸ı

-0.87

Frames

-0.80

ï¸

-0.71

Transform

-0.65

pload

-0.63

ersen

-0.62

rawdownloadcloneembedreportprint

-0.62

Shell

-0.62

axis

-0.62

ipel

-0.62

POSITIVE LOGITS

aucuses

0.81

 trooper

0.74

sylvania

0.72

hyde

0.71

 troopers

0.70

 reserve

0.65

 Marshal

0.62

itol

0.61

 Watkins

0.61

 legislatures

0.61

Activations Density 0.190%

government-related terms like titles, positions, and institutions

references to governmental or state-related entities and processes

No Comments

No Known Activations