INDEX

Explanations

critiques or discussions about whether systems are fair or rigged

oai_token-act-pair · gpt-3.5-turbo

topics related to critiques of societal systems

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 10-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.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

 Candle

-0.76

 sighting

-0.72

 Outdoor

-0.71

RTX

-0.71

 sightings

-0.70

 Laser

-0.70

 sqor

-0.69

 Bass

-0.68

 Fury

-0.68

 aloud

-0.67

POSITIVE LOGITS

 rigged

1.24

 functioning

1.22

 malfunction

1.17

 dysfunctional

1.09

 imperfect

1.07

 safeguards

1.04

 inefficient

1.03

 operated

1.02

 dysfunction

1.00

 corrupt

1.00

Activations Density 0.386%

critiques or discussions about whether systems are fair or rigged

topics related to critiques of societal systems

No Comments

No Known Activations