INDEX

Explanations

The main thing this neuron does is find references to human resources or staffing

oai_token-act-pair · gpt-3.5-turbo

references to personnel 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 @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.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

rous

-0.80

Brow

-0.77

lier

-0.69

Mother

-0.68

PUT

-0.66

Deal

-0.65

ebook

-0.65

wit

-0.65

Pic

-0.64

fin

-0.63

POSITIVE LOGITS

 personnel

1.27

andowski

1.00

 Personnel

0.97

 veter

0.95

chwitz

0.95

irtual

0.94

ynski

0.94

ilitary

0.89

 corps

0.88

 morale

0.84

Activations Density 0.006%

The main thing this neuron does is find references to human resources or staffing

references to personnel in various contexts

No Comments

No Known Activations