INDEX

Explanations

The neuron is looking for terms related to monetary or numerical values

oai_token-act-pair · gpt-3.5-turbo

words and abbreviations related to monitoring or measurements

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 0-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.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.0.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

iste

-0.96

ice

-0.79

amental

-0.76

nel

-0.74

ide

-0.74

ices

-0.73

ine

-0.73

otted

-0.73

oret

-0.72

iston

-0.72

POSITIVE LOGITS

PER

2.38

HER

2.13

CAR

2.12

TIME

2.12

MON

2.08

CON

2.08

CHAR

2.07

DAY

2.05

MET

2.04

DES

2.03

Activations Density 0.053%

The neuron is looking for terms related to monetary or numerical values

words and abbreviations related to monitoring or measurements

No Comments

No Known Activations