INDEX

Explanations

names or terms related to individuals, possibly linked to news, broadcasting, or personal life

oai_token-act-pair · gpt-3.5-turbo

occurrences of the letter "K" 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 @ 6-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.6.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.6.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

terday

-0.77

ãĥ¼ãĥĨãĤ£

-0.75

 Marketable

-0.71

 acknow

-0.65

 contr

-0.63

Ø©

-0.62

theless

-0.61

ãĥ¼ãĥĨ

-0.61

 Celest

-0.60

 expires

-0.60

POSITIVE LOGITS

EEP

1.12

essel

1.11

oop

1.05

won

1.05

ratom

1.03

avan

1.02

ota

1.02

itty

1.01

oko

1.01

rieg

1.00

Activations Density 0.039%

names or terms related to individuals, possibly linked to news, broadcasting, or personal life

occurrences of the letter "K" in various contexts

No Comments

No Known Activations