INDEX

Explanations

terms related to technology and historical figures

oai_token-act-pair · gpt-3.5-turbo

elements related to specific objects, artifacts, or notable characteristics

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

 ASAP

-0.77

 upscale

-0.76

 Comcast

-0.73

 Playoffs

-0.71

 Spotlight

-0.70

 anyways

-0.69

DOJ

-0.69

 Nationwide

-0.68

 Deadline

-0.68

 paycheck

-0.68

POSITIVE LOGITS

igraph

0.98

stones

0.97

houses

0.97

 poems

0.97

 paintings

0.96

 deities

0.95

names

0.94

 motif

0.94

 deity

0.94

ograph

0.94

Activations Density 0.553%

terms related to technology and historical figures

elements related to specific objects, artifacts, or notable characteristics

No Known Activations