INDEX

Explanations

names of people or characters

oai_token-act-pair · gpt-3.5-turbo

names of people and their affiliations or roles 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 @ 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

Negative Logits

Statement

-0.82

 Pastebin

-0.80

 Defendant

-0.77

 Defendants

-0.72

 Forsaken

-0.70

 Official

-0.67

Appearances

-0.65

pite

-0.64

 Plaintiff

-0.64

 Mandatory

-0.63

POSITIVE LOGITS

 explains

1.24

 remembers

1.23

 recalls

1.17

 thinks

1.16

 recommends

1.14

 resear

1.11

 says

1.11

 agrees

1.10

 calculates

1.10

 teaches

1.06

Activations Density 0.323%

names of people or characters

names of people and their affiliations or roles in various contexts

No Known Activations