INDEX

Explanations

No Explanations Found

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

reci

-0.90

oqu

-0.73

Reason

-0.73

import

-0.72

usage

-0.72

mable

-0.68

urable

-0.67

 Quantity

-0.67

 Collector

-0.66

obar

-0.65

POSITIVE LOGITS

hea

0.80

wee

0.72

 tentacles

0.71

 Bailey

0.65

Ake

0.64

 heightened

0.63

 Kraken

0.61

 waning

0.60

 credibility

0.60

aleigh

0.59

Activations Density 0.000%

No Known Activations

This feature has no known activations.