INDEX

Explanations

numeric values related to counting, measuring, or quantifying

oai_token-act-pair · gpt-3.5-turbo

various forms of the word "total" and its derivatives related to amounts or counts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.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

nels

-0.77

gets

-0.75

edIn

-0.75

edin

-0.73

 injecting

-0.71

house

-0.71

 Handler

-0.66

eering

-0.66

uay

-0.66

culosis

-0.66

POSITIVE LOGITS

 totaled

0.87

 totals

0.84

 Amph

0.80

onies

0.79

 amphib

0.79

ESA

0.77

rophic

0.75

 totaling

0.75

pha

0.74

eneg

0.72

Activations Density 0.019%

numeric values related to counting, measuring, or quantifying

various forms of the word "total" and its derivatives related to amounts or counts

No Comments

No Known Activations