INDEX

Explanations

phrases related to granting or allowing access to something

oai_token-act-pair · gpt-3.5-turbo

instances of the word "access."

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

okia

-0.79

oche

-0.72

cz

-0.70

 depended

-0.69

NR

-0.68

 revolves

-0.68

need

-0.68

atform

-0.67

akura

-0.66

rors

-0.65

POSITIVE LOGITS

 Flavoring

0.77

 defend

0.69

 assist

0.68

 shore

0.67

 preserve

0.66

 lift

0.66

site

0.64

 Meridian

0.62

è¡

0.61

 pass

0.59

Activations Density 0.252%

phrases related to granting or allowing access to something

instances of the word "access."

No Comments

No Known Activations