INDEX

Explanations

names of universities

oai_token-act-pair · gpt-3.5-turbo

references to specific individuals and institutions, particularly related to Jim Crane and Cornell University

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.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.5.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

ktop

-0.85

tered

-0.73

 spare

-0.73

ternity

-0.70

alions

-0.70

nda

-0.69

utic

-0.68

 regimes

-0.68

 Hyundai

-0.66

ghan

-0.66

POSITIVE LOGITS

issance

0.84

 Borough

0.80

Bridge

0.72

hots

0.71

 Falls

0.69

inary

0.67

Ó

0.66

 Marble

0.65

River

0.64

LOD

0.64

Activations Density 0.044%

names of universities

references to specific individuals and institutions, particularly related to Jim Crane and Cornell University

No Comments

No Known Activations