INDEX

Explanations

instances of people being released from a situation or place

oai_token-act-pair · gpt-3.5-turbo

instances of the word "released" related to individuals' legal situations

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 4-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.4.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.4.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

eries

-0.69

stal

-0.68

iop

-0.68

Ì¶

-0.66

SL

-0.63

advertisement

-0.62

 Evolution

-0.62

gly

-0.61

ential

-0.61

akia

-0.61

POSITIVE LOGITS

 prisoners

0.94

 captives

0.84

 prisoner

0.83

released

0.79

 detainees

0.79

æĸ¹

0.79

ileaks

0.75

 hostages

0.75

parency

0.75

 releasing

0.73

Activations Density 0.029%

instances of people being released from a situation or place

instances of the word "released" related to individuals' legal situations

No Comments

No Known Activations