INDEX

Explanations

adjectives describing the intensity of a situation or experience

oai_token-act-pair · gpt-3.5-turbo

declarations of events or states that emphasize "it was" in various contexts

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

Negative Logits

soever

-0.89

 ILCS

-0.70

doms

-0.69

ilings

-0.68

raper

-0.67

lie

-0.67

ature

-0.65

ranch

-0.65

 accuser

-0.65

 Critics

-0.62

POSITIVE LOGITS

 kinda

1.19

 nice

1.09

 raining

1.08

 definitely

1.06

 amazing

1.05

 pretty

1.02

 funny

1.02

 weird

1.00

 awesome

1.00

 surreal

0.98

Activations Density 0.337%

adjectives describing the intensity of a situation or experience

declarations of events or states that emphasize "it was" in various contexts

No Known Activations