INDEX

Explanations

mentions of the word "am" followed by a number indicating intensity or degree

oai_token-act-pair · gpt-3.5-turbo

occurrences of the substring "am," which may indicate a focus on verbs or related concepts

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 kinderg

-0.67

 strawberries

-0.60

 affairs

-0.60

 puberty

-0.60

uper

-0.60

 wool

-0.59

 resemb

-0.58

yi

-0.58

 stakes

-0.58

lift

-0.57

POSITIVE LOGITS

endment

1.28

nesty

1.25

sterdam

1.16

bitious

1.15

ilies

1.13

pling

1.11

essage

1.06

azing

1.05

pering

1.05

munition

1.04

Activations Density 0.048%

mentions of the word "am" followed by a number indicating intensity or degree

occurrences of the substring "am," which may indicate a focus on verbs or related concepts

No Comments

No Known Activations