INDEX

Explanations

proper nouns, specifically names like "Philip"

oai_token-act-pair · gpt-3.5-turbo

occurrences of the name "Philip."

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

inventoryQuantity

-0.80

headers

-0.73

Ranked

-0.68

imilation

-0.68

ãĤ´ãĥ³

-0.67

available

-0.65

REAM

-0.65

maid

-0.65

yrinth

-0.64

LOAD

-0.64

POSITIVE LOGITS

 Philip

1.01

anthrop

0.98

osate

0.86

 Randolph

0.83

ronics

0.77

avorite

0.77

oshenko

0.77

obar

0.76

ython

0.75

 Morris

0.75

Activations Density 0.004%

proper nouns, specifically names like "Philip"

occurrences of the name "Philip."

No Comments

No Known Activations