INDEX

Explanations

names, titles, or labels starting with "Prim"

oai_token-act-pair · gpt-3.5-turbo

references to a specific term, "Prim," and its variations in context

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

lihood

-0.75

IGH

-0.73

äºº

-0.67

 Hanson

-0.65

jriwal

-0.65

kson

-0.64

hiro

-0.63

hold

-0.61

 Bard

-0.61

 Cheong

-0.59

POSITIVE LOGITS

itives

1.53

ordial

1.43

itive

1.42

etime

1.28

arily

1.23

rose

1.18

acy

1.07

aries

1.01

itiveness

0.99

ework

0.92

Activations Density 0.025%

names, titles, or labels starting with "Prim"

references to a specific term, "Prim," and its variations in context

No Comments

No Known Activations