INDEX

Explanations

instances where people or entities are mentioned together in relation to a topic or issue

oai_token-act-pair · gpt-3.5-turbo

conjunctions and words that suggest partnership or connection between entities

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 3-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.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

ISION

-0.67

 Magnus

-0.65

 qualifier

-0.65

 Prol

-0.63

 sunset

-0.62

herence

-0.61

lift

-0.59

Installation

-0.58

ãĤĵ

-0.58

 prol

-0.58

POSITIVE LOGITS

doms

0.97

 alike

0.87

 respectively

0.86

noxious

0.80

etc

0.74

pherd

0.72

vous

0.71

ruciating

0.69

 beware

0.67

who

0.66

Activations Density 1.075%

instances where people or entities are mentioned together in relation to a topic or issue

conjunctions and words that suggest partnership or connection between entities

No Comments

No Known Activations