INDEX

Explanations

The neuron looks for words related to facial features or functions

oai_token-act-pair · gpt-3.5-turbo

terms related to facial features or expressions

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.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.5.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

 Breed

-0.77

 Slug

-0.74

 extremes

-0.70

 Barkley

-0.68

hood

-0.68

Trend

-0.68

 Seas

-0.67

uyomi

-0.67

UCK

-0.67

 Sawyer

-0.67

POSITIVE LOGITS

ilitation

1.32

eless

1.11

ilit

1.08

ulty

1.02

terior

1.00

ep

0.97

ility

0.96

sim

0.94

ilities

0.94

iliate

0.94

Activations Density 0.019%

The neuron looks for words related to facial features or functions

terms related to facial features or expressions

No Comments

No Known Activations