INDEX

Explanations

names containing the letters "za"

oai_token-act-pair · gpt-3.5-turbo

references to specific individuals, particularly those with the name "Za."

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.8.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

lessly

-0.94

LESS

-0.72

ician

-0.68

lessness

-0.68

mingham

-0.68

gement

-0.67

ativity

-0.66

jriwal

-0.64

folk

-0.64

landish

-0.64

POSITIVE LOGITS

 Motorsport

0.91

emon

0.84

ql

0.82

Ga

0.81

Ti

0.80

eta

0.79

ombies

0.78

Corp

0.78

ombie

0.78

illi

0.77

Activations Density 0.057%

names containing the letters "za"

references to specific individuals, particularly those with the name "Za."

No Comments

No Known Activations